OpenAI Accuses Chinese AI Startup DeepSeek of Bypassing Safeguards to Replicate US Models

OpenAI has leveled serious allegations against Chinese AI firm DeepSeek, claiming the startup circumvented access restrictions to replicate and distill outputs from US-developed AI models. In a memo to the US House Select Committee, OpenAI detailed how DeepSeek employees allegedly used obfuscated third-party routers to access proprietary systems—raising concerns about AI security and technological integrity.

The dispute highlights escalating tensions in the US-China AI race. OpenAI warned that DeepSeek's actions risk creating AI systems without proper governance or consumer protections. The memo also cited political bias and censorship issues in DeepSeek's models, particularly around sensitive topics.

This confrontation underscores the broader battle over AI supremacy, where intellectual property safeguards are increasingly becoming geopolitical flashpoints. The allegations come as both nations vie for dominance in next-generation artificial intelligence.